Summary Along with chromosomal re-arrangement, copy number alterations, and epigenetic changes in the human genome, the acquirement and accumulation of somatic mutations are key determinants of cancer development. With the rapid progress of cancer genome studies, many somatic mutations in population of cancer cells have been identified. However, cancer related mutations are heterogeneous and may have moderate to low recurrency. It is difficult to assess if a particular mutation is related to tumorigenesis. In addition, it is challenging to detect significant effects of accumulated multiple mutations that are of low recurrency. The goal of this project is to develop computational tools and databases to aid in assessment of the roles of somatic mutations, in discovery of additional novel cancer variants, and in identification of their higher order cooperative effects. We will collect a significant portion of known somatically acquired variants in the coding exons of cancer cells and will map them to our precomputed 3D functional surface maps of protein structures. We will develop computational capabilities to enable understanding of the mechanistic roles of these variants by integrating results from analysis of evolutionary substitution patterns, changes in biophysical properties, and potential roles in biochemical pathways. Our computational tools will help to discover novel cancer variants and their higher-order cooperativities. The outcomes of this project will be: 1) a comprehensive database of surface maps of cancer related genes and their variants, with additional evolutionary pattern and biophysical properties computed for readily interpretable mechanistic insight, and 2) computational tools to aid in discovery of novel cancer variants and their higher order cooperative effects.